Commits · master · Vision / GenVQA · GitLab

Snippets Groups Projects

Sep 13, 2016
- change .gitmodules · 3af697f3
  tgupta6 authored 8 years ago
  
  3af697f3
Sep 04, 2016
- constants_vision_gpu_2 and train_no_pretraining · c32cb279
  tgupta6 authored 8 years ago
  
  c32cb279
- change in constants_vision_gpu_1 and something in train · 383d8d57
  tgupta6 authored 8 years ago
  
  383d8d57
- answer train uses qids for genome · cc38d8bd
  tgupta6 authored 8 years ago
  
  cc38d8bd
- add train_subset in eval · 4fa768d7
  tgupta6 authored 8 years ago
  
  4fa768d7
- obj atr eval uses new genome split · 319b70d3
  tgupta6 authored 8 years ago
  
  319b70d3
- select best model code for object attribute · b4e42ac1
  tgupta6 authored 8 years ago
  
  b4e42ac1
Sep 02, 2016
- changes in constants_crunchy · d3235a8e
  tgupta6 authored 8 years ago
  
  d3235a8e
Sep 01, 2016
- split genome dataset to avoid overlap with test set of vqa · f24081b8
  tgupta6 authored 8 years ago
  
  f24081b8
Aug 29, 2016
- Merge remote-tracking branch 'origin/multi_rate_training' · 895125c1
  tgupta6 authored 8 years ago
  
  895125c1
- compute_AP uses constants file · b918ff2a
  tgupta6 authored 8 years ago
  
  b918ff2a
- change in constants · c5f46bb7
  tgupta6 authored 8 years ago
  
  c5f46bb7
- minor constant changes here and there · 8980e4f7
  tgupta6 authored 8 years ago
  
  8980e4f7
- eval selected categories and begin to split genome using vqa · 01dbf0ef
  tgupta6 authored 8 years ago
  
  01dbf0ef
Aug 26, 2016
- pending vqa_eval changes · 5449c1bc
  tgupta6 authored 8 years ago
  
  5449c1bc
- updated constants_vision_gpu_2 · d97f23e7
  tgupta6 authored 8 years ago
  
  d97f23e7
Aug 25, 2016
- vqa evaluation code and answer classifier scripts changed to using qids · d3b9aaf6
  tgupta6 authored 8 years ago
  
  d3b9aaf6
Aug 24, 2016
- added vqa evaluation submodule · 0a260f40
  tgupta6 authored 8 years ago
  
  0a260f40
Aug 20, 2016
- Circular batch and train held out · d80c835a
  tgupta6 authored 8 years ago
  
  d80c835a
Aug 19, 2016
- local change in crunchy constants and vqa eval code · 8922eafe
  tgupta6 authored 8 years ago
  
  8922eafe
Aug 15, 2016
- split val into val_subset and val_rest · ad70b50a
  tgupta6 authored 8 years ago
  
  ad70b50a
- Merge branch 'master' of gitlab-beta.engr.illinois.edu:Vision/GenVQA · b5ad2ae0
  tgupta6 authored 8 years ago
  
  b5ad2ae0
- make data reader single threaded · 4aa6f545
  tgupta6 authored 8 years ago
  
  4aa6f545
- visualization code · 3bfdbb2f
  tgupta6 authored 8 years ago
  
  3bfdbb2f
Aug 11, 2016
- Merge branch 'master' of gitlab-beta.engr.illinois.edu:Vision/GenVQA · 1299edce
  tgupta6 authored 8 years ago
  
  1299edce
- select best model passes obj_atr_loss_wt to graph · aa27e765
  tgupta6 authored 8 years ago
  
  aa27e765
Aug 08, 2016
- param change in constants_vision_gpu_2 · d016894c
  tgupta6 authored 8 years ago
  
  d016894c
- Merge branch 'master' of gitlab-beta.engr.illinois.edu:Vision/GenVQA · 05917417
  tgupta6 authored 8 years ago
  
  Add obj_atr_loss_wt to constants file
  05917417
- petty change to constants_vision_gpu_2 · ebb601ac
  tgupta6 authored 8 years ago
  
  ebb601ac
- obj atr wt in constants file · 5093204d
  tgupta6 authored 8 years ago
  
  5093204d
- select best model · ee037267
  tgupta6 authored 8 years ago
  
  ee037267
- merge wo_joint_training and large_batch_size · f2defe74
  tgupta6 authored 8 years ago
  
  f2defe74
- prep to merge · 4c434ff5
  tgupta6 authored 8 years ago
  
  4c434ff5
- dump json for eval · ce265cdf
  tgupta6 authored 8 years ago
  
  ce265cdf
Aug 05, 2016
- add no pretraining qa training script · 28db477b
  tgupta6 authored 8 years ago
  
  28db477b
- add separate constants for different machines · fdf8d455
  tgupta6 authored 8 years ago
  
  fdf8d455
Aug 04, 2016
- before setup on gpu machines · 236ecc99
  tgupta6 authored 8 years ago
  
  236ecc99
Aug 03, 2016
- large batch size and cached features with pretraining code with cached features · bc4cf1ca
  tgupta6 authored 8 years ago
  
  bc4cf1ca
Jul 26, 2016
- working batch size 1 with cached features · 6ad42fc4
  tgupta6 authored 8 years ago
  
  6ad42fc4
Jul 24, 2016
- add code to cache resnet features · 8a23bd6d
  tgupta6 authored 8 years ago
  
  8a23bd6d